PRELIMINARY RESULTS OF CLIMATE/COTTON RAT ANALYSIS AND SUGGESTIONS FOR CLIMATIC MODELING PROCEDURES
Monthly data of Sigmodon hispidus (cotton rat) and Peromyscus maniculatus (deer mouse) populations were collected from September, 1973 to December, 1990 at the Nelson Environmental Study Area in Lawrence, Kansas. These were made available to the Center for Climatic Research at the University of Delaware for preliminary climatological analysis. The goal of the preliminary study is to understand the population dynamics of the rodents and to determine crude climatological relationships for the purposes of developing sophisticated climatological modeling procedures which will be utilized within this cooperative agreement. The ultimate objective is to understand S. hispidus and P. maniculatus/climate interactions for the purposes of developing predictive equations for numerous locales around the country. The discussion here will concentrate on the preliminary analyses developed for S. hispidus; the P. maniculatus analyses are under way.
It appears that both rodent species exhibit large seasonal variations in population (Figure 1). For S. hispidus, a strong seasonal maximum is reached in fall, and a minimum occurs in March and April. The seasonal maximum and minimum for P. maniculatus is somewhat different, and occurs in winter and late summer, respectively. An analysis of monthly temperature and precipitation values indicates inverse relationships between both of these climate variables and populations of the rodents.
The cotton rat population seems to be affected both by extremes in temperature and precipitation. A rapid decline in winter is especially noted when the weather is particularly cold. For example, during the cold winter of 1983-84, when the mean temperatures for December and January were 15oF and 26oF, respectively, the cotton rat population plummeted to 0, and no individuals were captured again until August, 1984. In addition, the cotton rat population appeared to become nearly locally extinct from June, 1975 to March, 1977. It appears that a highly variable precipitation regime was responsible for this population decrease, especially from midsummer, 1975 through the summer of 1976. For example, July, 1975 was very dry (precipitation of less than one inch), September was wet (over four inches), October was exceedingly dry (virtually no rain), and November was wet (almost five inches). A very dry winter followed, along with a wet spring in 1976. According to Swihart and Slade (1990), inconsistent precipitation during this period could limit the nutritional value of vegetation consumed by cotton rats, thus stressing the population. Interestingly, while the cotton rat population was very low during this period, the deer mouse population was quite high. An apparent inverse relationship between populations of S. hispidus and P. maniculatus has been noted, and the competition between these two species requires further analysis.
Relationships between mean monthly precipitation and temperature and mean monthly rodent populations have been developed, and these have proven to be statistically significant. For both the cotton rat and deer mouse, inverse relationships exist between temperature, precipitation, and rodent populations. These relationships become stronger when the precipitation and temperature values are lagged (e.g. the relationship between October cotton rat populations and September precipitation; this represents a one month lag). For example, cotton rat populations seem to respond very strongly to three month temperature and precipitation lags; the R2 is 0.74 and 0.53, respectively (3 month lag: Figure 2). In addition, for the three month lag, the relationships are now direct. Considering that these relationships have been established for monthly means, it is quite likely that a more sophisticated climatological analysis, using the actual monthly data over the period of record, will yield more revealing results (Temp.: Figure 3);(Prec.: Figure 4). The impact of specific stressful periods will be noted, and threshold conditions which lead to rapid increases and decreases in organism populations, will be established.
The results of these preliminary analyses have provided significant guidance involving the climatic modeling necessary in the cooperative agreement. There is no doubt that a complete water budget analysis is required to understand surface-atmosphere energy/moisture processes which lead to stress for both the cotton rat populations and the nutritional value of the vegetation it consumes. As described previously, climatic water budgets describe evapotranspiration rates, moisture surpluses, and moisture deficits, all of which have been shown to stress a large variety of organism populations when values are very low or very high. The water budget development requires climate data input which is rather modest, and is collected at cooperative weather stations around the country. In addition, it permits the development of quantitative relationships between energy/moisture variables and rodent populations, which can be utilized for predictive purposes. Most previous studies of this type rely on raw weather variables such as temperature and precipitation. These parameters do not describe adequately the actual energy/moisture balance of a given rodent habitat. Thus, the water budget will be employed to define climate limiting factors more precisely which are no doubt strongly influential on the population dynamics of the cotton rat and deer mouse.
The use of air mass-based synoptic methodologies will also provide needed climatic information in this study. As mentioned earlier, air mass-based analyses describe the holistic impact of atmospheric factors on organisms, and permit simultaneous evaluations of numerous weather elements (as they realistically exist) upon populations. Each day can be categorized statistically within an air mass type, and those air masses which contribute to heightened or diminished populations can be determined. Air mass analysis has been used successfully to describe climate limiting factors for humans (Kalkstein, 1991) and numerous other organisms (Yarnal, 1993), and it is expected that the population dynamics of these rodents are also controlled by certain air masses which may be beyond the threshold of their tolerance.
Deer mouse pop. (1 month lag) vs. T, Precip.
Return to the SCL Home Page